Month: October 2019

Sellers.json was created to increase visibility of how impressions are resold and by whom. That would help reassure buyers that they’re not being duped by shady players who aren’t authorized to resell a publisher’s inventory. But until two months ago, adoption of the tool was low, according to the five ad tech execs interviewed for this article.

Two months ago, demand-side platform The Trade Desk changed that when it gave ad exchanges an ultimatum: Use the sellers.json tool to show buyers whether they sell inventory directly from a publisher or a reseller, or get dropped as a partner and miss out on millions of dollars worth of ad spend the DSP bids in online auctions.

“The days of taking an impression and passing it from one SSP to another SSP, to another SSP and then to a DSP are over,” said Dave Pickles, chief technology officer and co-founder of The Trade Desk.

A total 123 exchanges, including 34 of the 50 most widely adopted, have shared a sellers.json file as of Sept. 28., according to programmatic consultancy Jounce Media. Google, Freewheel and Teads are the three most notable ad tech players not to have posted a sellers.json file yet. Teads will share the fie in November, said Eric Shih, global svp of business development, while both Google and Facebook have made private commitments to do the same by the end of the year, confirmed two ad tech execs on condition of anonymity.

Previously, if an advertiser wanted to check whether an exchange directly worked with a publisher or not, they needed a way to pull reports directly from those ad tech vendors. But ad exchanges had no reason to share that information, and media buyers had no way to extract that information from them. The impasse made it hard to spot duplicate bid requests from the same exchange for the same impressions. Roughly 25% of the 250 billion bid requests translated daily by Bidswitch are actually unique, for example. Simply put, the number of bid queries processed daily vastly exceeds the number of possible ad impressions that can actually be viewed by human eyeballs.

“The widespread adoption of the files over the last few weeks is giving us a much richer picture behind each of these supply paths,” said Chris Kane, founder of Jounce Media. “We’ve had no way of knowing why an exchange has five different integrations with the same publisher until now.”

Some SSPs like Triplelift are going so far as to use sellers.json to teach buyers how to find those exchanges that add value.

“If you were an exchange of a certain size that couldn’t scale its exclusive relationships you had with publishers then they were considered a liability,” said Ari Lewine, chief strategy officer of ad exchange Triplelift. “Now, those relationships will become the most prized possession on the supply side.”

But that won’t necessarily mean that all resellers face the chop. Some resellers provide unique placements on publisher sites that aren’t available elsewhere, while others offer unique execution of widely available ad formats, along with those that offer unique targeting.

“It’s not necessarily the case that sellers.json will cut a bunch of resellers,” said Tom Kershaw, chief technology officer at The Rubicon Project. “It’s going to require that anyone who wants to be in the supply path do something of value. Just do something other than nothing. That’s a reasonable request.”

The post To reduce auction duplication, buyers start to enforce sellers.json appeared first on Digiday.

“Longer-term, I think this is a necessary step as Facebook consolidates the Messenger ecosystem across all their platforms and opens up further advertising opportunities there,” he said.

Three years ago, Messenger drew plenty of interest from brands who saw it as a tool for customer service, commerce, discovery and entertainment, with some even building chatbots specific to the platform. But their interest dampened when it was revealed that Facebook’s AI-powered bots could only respond to 30% of requests without some sort of human intervention.

But for the right types of brands and use cases, Facebook Messenger can still be meaningful for the more than 140 million different businesses using it and Facebook’s other platforms.

The new feature, in particular, solves a challenge companies have had with doing business on chat platforms: It invites brands easily and directly to engage with consumers. “It’s hard for brands to earn permission to show up in that space and not just be immediately rejected,” Mike Dossett, vp and director of digital strategy for RPA.

While brands’ overall usage of Messenger may have dipped over the years, said Huynh, it’s more so because they better understand its value for specific types of brands and use cases.

This latest feature is especially useful for lead generation, said Gurbani Chadha, associate director of strategy, digital for Deutsch Los Angeles. “It’s a great tool to direct people to that one-to-one messaging setting, and to create conversions. It’s giving people something back for engaging with a brand.”

Transportation company Cardinal Logistics, for example, used Messenger to recruit new drivers, and saw a nearly two-times increase in quality leads while reducing their costs by 55%.

Other businesses are using Messenger as a customer service platform. Chadha said her agency’s clients see many of those service-related conversations taking place with Gen Xers and Boomers.

“We always tell our clients that if they want to use Messenger, or to use chatbots, especially, it needs to be more efficient than other available alternatives,” said Dossett. “If it solves a pain point, then use it.”

“We find the best messaging experiences are a balance of automation and human agents,” said Ze’ev Rosenstein, Facebook’s head of people to business messaging for Messenger. “Automation can easily handle certain types of interactions, such as greetings, serving up FAQs, or tracking a package, while freeing up human agents to handle more complex inquiries.”

Canadian airline WestJet saw customer sentiment increase nearly 25% since it launched its own Messenger chatbot. Employee productivity has also improved because Juliet now oversees 50% of all customer inquiries and can take five times more messages than traditional customer service channels.

Another popular use case, they noted, is for commerce. Burger King, for instance, enables people to create a meal and pay for it through Facebook Messenger, and then find a local Burger King location where they can pick up their meal.

In 2018, fashion company Michael Kors developed a Facebook Messenger chatbot that enabled people to find specific items and also answered any questions they had. It has since garnered more than 375,00 active users for the brand, and an average of 45,000 new users per month. The average containment rate on the chatbot is 98.32%, meaning that almost all who use the chatbot are staying with it as opposed to being directed to a customer service agent.

Last year, Nike did an exclusive Kyrie 4 sneaker drop via Facebook Messenger, and sold out of its inventory in less than an hour.

QVC has also used Messenger to create sponsored messages that generated 5.3 times higher return on ad spend compared to other digital solutions.

One thing marketers wish Facebook would change about Messenger, however, is that people who don’t have the Messenger app downloaded separately can’t benefit from a product feature like this because they’re directed to a dead end, Chadha noted. Keeping people in the same platform would be an even bigger benefit to marketers, she said.

Photo credit: Facebook

The post Is Facebook Messenger still a viable channel for brands? appeared first on Digiday.

Some publishers will have some advantages when that switch happens. And a number of key questions, such as how advertisers balance between host-read and network-optimized ads, or how platforms will monetize shows on their platform, remain. But as podcasting speeds toward becoming a billion-dollar market by 2021, up from the $479 million generated in 2018, that growth will come from brand advertisers, who demand scale and ease from sellers.

That means podcasting will inevitably become more like the display ad ecosystem: commoditized, automated, audience-driven rather than contextual. The podcast ad world has been craft; it’s destined to be mechanized.

“It’s really hard to scale the program-specific, custom host-read stuff,” said an executive at one media agency that invests in audio. “It used to be death by a thousand cuts. … If you were interested in doing podcasting as a client, and you could work with two or three big partners, that is much more interesting than trying to do a hundred different deals.”

The balance of power has not tipped yet, in part because the biggest platforms are still getting their ducks in a row; many of the acquisitions that brought podcasting onto platforms including Spotify and Pandora were made less than a year ago.

“I think they’re still learning about what they bought,” said Stephen Smyk, svp of podcasting and influencer spending at the media agency Veritone One.

But the rest of the industry is trying to get itself ready for a world where marketers expect mass reach and compliance with industry metrics. Late last week, Stitcher’s ad network, Midroll, gave shows in its network a timeline for moving over to a different hosting solution, as part of a gradual shift toward an impression-based system that would allow Midroll to sell advertising differently.

“The rubber’s hitting the road,” Stitcher CEO Erik Diehn said. “There’s so much inventory out there right now that’s inefficiently transacted when you do it on an episode-by-episode basis.”

That efficiency will allow more brand dollars to pool into the space. Brand advertising now constitutes 38% of podcast revenue, according to IAB figures; branded content accounts for another 10%. Those buyers are accustomed to simplicity and scale, two things that buyers say podcasting, to date, has struggled to deliver.

Other players in the industry are acting accordingly. In August, Art19, the enterprise podcast hosting and monetization platform, announced it had launched an ad network, which would allow buyers to place ads on a listener-by-listener basis across 200 different shows.

While individual publishers will never be able to deliver the scale that platforms offer, some hope to counter by rolling their podcasting into the multifaceted deals they are trying to construct for advertisers. Being able to construct an ad program that includes digital, print, audio and other formats could give publishers a head start in proving out how podcast ads fit into multi-touch attribution models, Soon said.

“What we’re seeing is these platforms really following where the dollars are,” said Zoe Soon, the IAB’s vp of mobile. “It’s on publishers to say, ‘We have this audience; here are the ways we can reach them.’”

Others, such as Barstool Sports, are planting a flag. Late Tuesday, ahead of Barstool’s first-ever upfront, CEO Erika Nardini uncorked a rant on Twitter about where podcast advertising was going, vowing that Barstool, which generates a third of its revenues from podcast ads, would never go full programmatic.

“When you make ads the skippable middle man, they will get skipped,” Barstool Sports CEO Erika Nardini wrote on Twitter Tuesday.  “When you are a part of only one piece but not the entire conversation, you will get lost.”

“This is unsurprising and media gentrification at its finest,” said Dan Granger, the CEO of Oxford Road. “Podcast started as a free-for-all, and those of us who were early enough into the space all benefited greatly. But now they are going to start imposing new rules and processes that dilute the value of every impression beneath what it was before.”

The post As platforms and ad tech circle, podcasting’s small, beautiful age is drawing to a close appeared first on Digiday.